Top
Best
New

Posted by JonSchneider 12 hours ago

Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint(prismml.com)
419 points | 125 commentspage 2
jedbrooke 10 hours ago|
Running at about 7-8 tok/s (~60 tok/s prefill) on a Mac Mini M2 16GB.

So far feels smarter than Bonsai 1 27B, it’s slightly larger than the Q1_0 quant. Super exciting stuff :)

kamranjon 12 hours ago||
Love this for the folks with 16gb graphics cards - 3.8 27b has been incredible but not quite runnable on anything less than 32gb - will try loading this up on my 16gb intel b50 and see how it goes - not sure these quants can be accelerated by the XPU cores yet but maybe in time!
kennywinker 4 hours ago||
I run qwen 27b on an old-ass 16gb gpu. It’s very possible using unsloth 2bit and 3bit quants, tho there are a bunch of interesting quants that let you run closer to 4bit on 16gb. This article that’s currently also on the front page mentions a bunch of them while discussing their own quant https://byteshape.com/blogs/Qwen3.8-27B/
kadoban 11 hours ago|||
You can run the ~4 bit quant(s) on 24gb, if you're not _too_ picky on context size.

This will hopefully be better, though it'd be a _very_ surprising increase in performace at the size they say. Would love to see more about how it benchmarks.

spijdar 11 hours ago|||
I run Unsloth's UD-Q4_K_S on 20 GB of VRAM (RX 7900 XT) and I get ~90k tokens of context without quantizing KV cache. With 8-bit quantization, I get about a 134k token context window. That's with only one slot, but for me, it works pretty darn well, with 20-35 tok/s depending on how full that window is.
orsorna 9 hours ago||
7900 XT is a sleeper card. When I initially bought it, it was priced at the lowest wattage per $ per GB VRAM (not normalized for token speeds...) Although I ended up swapping for the XTX because that 4GB means everything in just increasing the context window. At 8bit KV my window is over 200k, and although qwen3.8 loves vomiting out tokens as part of its reasoning chain I trust it enough to get assigned tasks done eventually, which I could not say of any model before its release.
kadoban 5 hours ago||
How has software/driver support been? I got burned hard by AMD last generation or the one before. Things smoother now, or do you have to baby it like hell and pick and choose software that works?
orsorna 5 hours ago||
I don't do anything fancier than inference, and I only use llama.cpp, which supports rOCM. I've had few issues; most GGUFs I download work right out of the box. Nearly any popular model has a quant that just works. But as you can see I don't use my GPU for anything weird or nonstandard.
redox99 6 hours ago||||
You can trivially run 131k on 24GB 4bit, and there are repos with tweaks that allow you to get the full 262k but idk if there's degradation with their approach.
lta 8 hours ago|||
I'm doing the same with a context of about 128-150k Surprisingly, I get subjectively better results with Unsloth's 3 bit quants (UD-Q3-XL something), than their 4 bit quants (S or M)
djkoolaide 8 hours ago||
Tried it today on a B70 and couldn't get anything usable out of it.

Prism's llama.cpp fork only has the kernels for CUDA, CPU and Vulkan. No SYCL at all :(

g023 8 hours ago||
They need to make a Big Bonsai, something at the enterprise levels that can compete with DSV4 Flash etc.
Fordec 8 hours ago|
A Tree if you will
ncr100 5 hours ago||
Or Grove
JonSchneider 11 hours ago||
I'm hoping they release an 8B v2 based on the Qwen 3.8 series in the near future - that would give us a really powerful model that could be run directly on users phones.
verdverm 10 hours ago||
That would require Alibaba releasing a Qwen 3.8 8B first
sroussey 11 hours ago||
Yes! And maybe get a hf fused webgpu runner for that model so it’s fast!
huseyinkeles 5 hours ago||
Testing on a MBP m4 pro 24gb

~100t/s prefill, ~15t/s, dropping to ~10t/s later with 64k context.

The issue is I have yet to find a useful agentic local llm that I can run on this machine.

Just given a relatively simple task on a swift app, took 25 minutes, brainstorming like crazy but can not decide on what to do. Eventually I killed it. GPT 5.6 sol-medium took 3 minutes to complete the same task for reference.

aetherspawn 5 hours ago||
Gemma 30B with 256K context runs at 20 tok/sec on my M3 Max with 128GB RAM so I think there’s something wrong with your setup. This should run at ~30-40 toks. Maybe your inference engine is not optimised for Mac.
huseyinkeles 5 hours ago|||
I just used their `Bonsai-demo` repo like this;

`cd ~/Code/Bonsai-demo && BONSAI_CTX=65536 ./scripts/start_llama_server.sh`

then used it in a very minimalistic pi with a very small system prompt.

Didn't spend much time to try to optimize it tbh, but my issue was not the speed. it just could not make a decision on how to implement the task, kept going on an on.

piyh 4 hours ago|||
30B is useless on 24 gigs of ram as there's ~4 gigs of ram left for everything else even with unsloth quants
sean_pedersen 5 hours ago||
Try a MoE model like Qwen3.6 35B-A3B for better tok/s
huseyinkeles 5 hours ago||
I tried this one, but I found Ornith1.5 to be a better MoE model for me, also very fast. But I still couldn't make it implement a real task on a real repo :( it only worked with an extremely clear directions and very small tasks.
mpweiher 1 hour ago||
Is it just me or are local models getting better (catching up) a lot faster than the frontier models are getting better (creating distance)?

If true, that would be a very welcome development.

redox99 7 hours ago||
I tried their WebGPU version and it immediately started looping. Yeah "near lossless" my ass. Plus the reasoning that it looped on was clearly wrong and unlike the non quantized 27B
circularfoyers 9 hours ago||
I wonder how their talks with Apple went. Having this run on the TPU opposed to just the GPU, which drains a significant amount of battery life by comparison, is what I'm really interested in.
Havoc 10 hours ago||
Cautiously optimistic. The V1 was noticeably weak on world knowledge but here the 3.8 base model is geared more towards reasoning than world knowledge anyway so might not matter as much
2001zhaozhao 11 hours ago|
I think if they made this for Qwen3.8-Next it could fit in a single 5090?
kennywinker 4 hours ago|
180b * 1.76 bits per weight = 39.6 gigabytes.

Best you could realistically run in 32gb is like 28gb, or a 127B param model

jokethrowaway 50 minutes ago||
Qwen3.8-Next, thanks to its new architecture, is quite fast even if part of it is streaming from disk
More comments...